Assignment 2 requires writing a report or critique on the paper that you chose in Assignment 1 to Presentation and Participation component above. Your report should be limited to approx. 1500 words...

2
Software Engineering In Machine Learning – A Case Study
SOFTWARE ENGINEERING IN MACHINE LEARNING:
A CASE STUDY
Table of Contents
1. Software Engineerin and Machine Learning –
An Introduction ------- 3
2. Machine Learning Definition ------- 4
3. ML Workflow -------- 5
4. Software Engineering principles
in Machine Learning -------- 6
5. Machine Language Stages -------- 7
6. Machine Learning and Artificial Intelligence ------ 8
7. Machine Learning Techniques -------- 9
8. ML Process Maturity model -------- 11
9. Limitations of Machine Learning ------- 12
10. Conclusion --------- 13
11. Bibliography --------- 13
Software Engineering and Machine Learning – An Introduction
Software Engineering in Machine Learning involves discovering new data, managing the data and creating various versions for the data. They are more complex and difficult than in other applications.Software teams use different skills for customization and reuse.Machine Learning has AI components that are “entangled” in a more complex way. Hence, they are difficult to handle than the traditional software components. The behavior is non-monotonic.
If we see the Software History, earlier we had Personal computing, then the Internet and then the Web. Mobile computing and Cloud computing are found presently. Thus there are updates everyday in the dominant applications in the Software domain. New softwares were created to address these domains.
Currently Artificial intelligence(AI) and ML are used in software industry market.Different approaches of reasoning, problem solving, planning and learning are used in AI in Machine Learning. Statistical Modeling is used in ML.
Bing Search and Cortana virtual assistant of Microsoft uses machine learning. Text, voice and video translator uses Machine language. They are also used in Cognitive vison, Cognitive speech, and language translators. The customers builds machine learning application using the Azure AI platform [2]. The pre existing components of AI are leveraged and new areas of expertise is created. The Agile software processes is updated with AI workflows.
Machine Learning - Definition
Microsoft combine pre-existing software engineering capabilitiess with AI.Software teams are using feedback-intense Agile methods to develop their software [1], [15], [17] . They address changing customer needs through faster development cycles.
The paper discusses the following:
1) A description of how several Microsoft software engineering teams work cast into a nine-stage workflow for integrating machine learning into application and platform development.
2) A set of best practices for building applications and platforms relying on machine learning.
3) A custom machine-learning process maturity model for assessing the progress of software teams towards excellence in building AI applications.
4) A discussion of three fundamental differences in how software engineering applies to machine-learning–centric components vs. previous application domains
Process changes not only alter the day-to-day development practices of a team, but also influence the roles that people play.
15 years ago, many teams at Microsoft relied heavily on development triads consisting of a program manager (requirements gathering and scheduling), a developer (programming), and a tester (testing) [6]. These teams’ adoption of DevOps combined the roles of developer and tester and integrated the roles of IT, operations, and diagnostics into the mainline software team.
In recent years, teams have increased their abilities to analyze diagnostics-based customer application behavior, prioritize bugs, estimate failure rates, and understand performance regressions through the addition of data scientists [7], [8], who helped pioneer the integration of statistical and machine learning workflows into software development processes.
ML Workflow
One commonly used machine learning workflow at Microsoft has been depicted in various forms across industry and research [2], [5], [14], [15]. It has commonalities with prior workflows defined in the context of data science and data mining, such as TDSP [19], KDD [13], and CRISP-DM [14]. Despite the minor differences, these representations have in common the data-centered essence of the process and the multiple feedback loops among the...